tyrosinekinase2-mediatedsignaltransductionint ... · a crystal structure of the pseudoki-nase...

Tyrosine Kinase 2-mediated Signal Transduction in TLymphocytes Is Blocked by Pharmacological Stabilization ofIts Pseudokinase Domain*□S

Received for publication, October 16, 2014, and in revised form, March 5, 2015 Published, JBC Papers in Press, March 11, 2015, DOI 10.1074/jbc.M114.619502

John S. Tokarski‡, Adriana Zupa-Fernandez§, Jeffrey A. Tredup¶, Kristen Pike�, ChiehYing Chang‡, Dianlin Xie¶,Lihong Cheng§, Donna Pedicord**, Jodi Muckelbauer‡, Stephen R. Johnson‡, Sophie Wu¶, Suzanne C. Edavettal¶,Yang Hong‡‡, Mark R. Witmer¶, Lisa L. Elkin�, Yuval Blat**, William J. Pitts‡‡, David S. Weinstein‡‡,and James R. Burke§1

From the Departments of ‡Molecular Structure and Design, §Immunosciences Biology, ¶Protein Science, **Leads Discovery andOptimization, and ‡‡Discovery Chemistry, Bristol-Myers Squibb Research and Development, Princeton, New Jersey 08543 and the�Department of Leads Discovery and Optimization, Bristol-Myers Squibb Research and Development, Wallingford, Connecticut 06492

Background: Interleukin-23 mediates pathobiology in many autoimmune disorders.Results: A chemogenomics approach identified small molecule agents that block receptor-mediated activation or tyrosinekinase 2 (Tyk2) and downstream signaling. Compounds stabilize the pseudokinase domain of Tyk2.Conclusion: Small molecule ligands of the Tyk2 pseudokinase domain stabilize an autoinhibitory interaction with the catalyticdomain.Significance: This work enables the discovery of selective therapeutics targeting Tyk2-dependent pathways critical inautoimmunity.

Inhibition of signal transduction downstream of the IL-23receptor represents an intriguing approach to the treatment ofautoimmunity. Using a chemogenomics approach marryingkinome-wide inhibitory profiles of a compound library with thecellular activity against an IL-23-stimulated transcriptionalresponse in T lymphocytes, a class of inhibitors was identifiedthat bind to and stabilize the pseudokinase domain of the Januskinase tyrosine kinase 2 (Tyk2), resulting in blockade of recep-tor-mediated activation of the adjacent catalytic domain. TheseTyk2 pseudokinase domain stabilizers were also shown toinhibit Tyk2-dependent signaling through the Type I interferonreceptor but not Tyk2-independent signaling and transcrip-tional cellular assays, including stimulation through the recep-tors for IL-2 (JAK1- and JAK3-dependent) and thrombopoietin(JAK2-dependent), demonstrating the high functional selec-tivity of this approach. A crystal structure of the pseudoki-nase domain liganded with a representative example showedthe compound bound to a site analogous to the ATP-bindingsite in catalytic kinases with features consistent with highligand selectivity. The results support a model where thepseudokinase domain regulates activation of the catalyticdomain by forming receptor-regulated inhibitory interac-tions. Tyk2 pseudokinase stabilizers, therefore, represent anovel approach to the design of potent and selective agentsfor the treatment of autoimmunity.

In several human autoimmune diseases, such as psoriasis,rheumatoid arthritis, Crohn’s disease, and multiple sclerosis, akey pathogenic role for T helper 17 (TH17)2 cells in mediatinginflammation and tissue injury has been shown (for a review,see Ref. 1). Targeting the expansion and action of pathogenicTH17 cells or mediators produced by these cells, therefore, hasgarnered considerable interest as a strategy toward the discov-ery of novel therapeutic agents. Particularly intensive effortshave been directed toward the discovery of agents that targetinterleukin-23 (IL-23), a cytokine critical in the expansion andsurvival of pathogenic TH17 cells as well as the induction ofinnate lymphoid cells in autoimmunity (2, 3). Blocking mono-clonal antibodies directed against either the p19 subunit ofIL-23 or the p40 subunit it shares with IL-12 are being investi-gated in autoimmune diseases with clinical benefit demon-strated in patients with psoriasis and Crohn’s disease (4, 5).

Small molecule therapeutics targeting the IL-23 receptor (IL-23R) pathway represent especially intriguing approaches toautoimmunity, not only because of target load limitations withanti-cytokine monoclonal antibody therapeutics (e.g. the needto deliver sufficient antibody in relation to the level of cytokine),but perhaps more importantly, small molecule therapeuticsprovide the opportunity to target tissues that would be difficultto target with large protein-based therapeutics. Indeed, anti-p40 monoclonal antibodies have failed to show benefit inpatients with multiple sclerosis, possibly due to an inability ofthe antibody to penetrate the central nervous system (6, 7).

* All authors are employees of Bristol-Myers Squibb Co.□S This article contains supplemental Table S1.1 To whom correspondence should be addressed: Immunosciences Biol-

ogy, Bristol-Myers Squibb Research and Development, Mail Stop K24-03, P.O. Box 4000, Princeton, NJ 08543. Tel.: 609-252-4000; E-mail:[email protected].

2 The abbreviations used are: TH17, T helper 17; IL-23R, IL-23 receptor; Tyk2,tyrosine kinase 2; PTFE, polytetrafluorethylene; TSA, thermal shift analysis;BisTris, 2-[bis(2-hydroxyethyl)amino]-2-(hydroxymethyl)propane-1,3-diol;TCEP, tris(2-carboxyethyl)phosphine; CAR, cell active rate; ITC, isothermaltitration calorimetry; SH2, Src homology 2; PK, pseudokinase; PDB, ProteinData Bank.

THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 290, NO. 17, pp. 11061–11074, April 24, 2015© 2015 by The American Society for Biochemistry and Molecular Biology, Inc. Published in the U.S.A.

APRIL 24, 2015 • VOLUME 290 • NUMBER 17 JOURNAL OF BIOLOGICAL CHEMISTRY 11061

by guest on Novem

ber 6, 2020http://w

ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

The Janus protein kinase (JAK) family members tyrosinekinase 2 (Tyk2) and JAK2 are critical in the signal transductionpathways downstream of the IL-23 receptor (8). However,potential therapeutic agents that inhibit JAK2 would also resultin anemia and other untoward hematopoietic defects due to itscritical role in receptor signaling downstream of the receptorsfor erythropoietin, thrombopoietin, and other growth factors(8, 9). Targeting Tyk2 represents a more intriguing approach,especially because Tyk2-deficient mice are protected frommany models of experimental autoimmunity, including colla-gen-induced arthritis and experimental autoimmune encepha-lomyelitis (10, 11). However, due to the high homology betweenthe active sites in JAK kinase catalytic domains, only modestprogress has been reported in the design and identification ofpotent and highly selective inhibitors of the catalytic activity ofTyk2 as potential therapeutics (12, 13).

In an effort to discover alternative and potentially more trac-table kinase targets (and lead inhibitors) in the signaling path-way downstream of IL-23R, the present report details a che-mogenomics approach that ultimately led to the identificationof potent and selective inhibitors. These molecules act on Tyk2but not by binding to the catalytic domain of the kinase andinhibiting catalytic activity, which is the classical mode for pro-tein kinase inhibitors. Instead, the compounds were shown toprevent the receptor-mediated activation of the Tyk2 catalyticdomain (also known as the JH1 domain) as a consequence ofcompound binding to and stabilizing the adjacent pseudoki-nase domain (also known as the JH2 domain), so called becauseit is evolutionarily related to protein kinases but is not thoughtto support catalytic activity (14). The pseudokinase domains ofJAK family kinases have previously been implicated to play anautoinhibitory role in regulating activation of the adjacent cat-alytic domains (15, 16), but the mechanism by which this occursis poorly understood. The cellular, biophysical, and structuralbiology studies described herein detail the mechanism by whichthese compounds block Tyk2-mediated signaling and tran-scription of the IL-23R receptor. This work provides the firstdemonstration that downstream signal transduction can beinhibited by pharmacological modulation at the level of thepseudokinase domain, and the findings also suggest that theregulation of Tyk2 activation may be unique to this JAK familymember. This discovery provides an especially promising andnovel approach to the design of potent and selective therapeu-tics for the treatment of autoimmune diseases dependent onTyk2 signaling pathways.

EXPERIMENTAL PROCEDURES

High Throughput IL-23-stimulated Transcriptional ResponseAssay—kit225 human T cells with stable integration of a fireflyluciferase reporter gene under the control of the interferon-�activation sequence (IRF1-GAS-Luc) were grown in RPMI con-taining L-glutamine, 10% FBS, 20 ng/ml recombinant IL-2(BIOSOURCE, PHC0023), and Geneticin at 0.7 mg/ml (from 50mg/ml stock Gibco-BRL, 10131-035). Prior to the assay, cellswere washed three times in assay medium (phenol red-freeRPMI, 10% heat-inactivated FBS, 1% penicillin/streptomycin)to remove IL-2 present in the growth medium and allowed toincubate overnight. Immediately prior to assay, test com-

pounds were dispensed into assay plates from library sourceplates via acoustic dispensing, followed by the addition of65,000 cells/well in a total volume of 30 �l (10 �M final com-pound; 0.5% DMSO). IL-23 (prepared in PBS with 0.1% BSA)was added at a final concentration of 0.02 �M and allowed toincubate for 5 h at 37 °C. After incubating, 25 �l of Bright-Glo(Promega Corp.) was added prior to imaging the luminescentsignal on the ViewLux.

For the primary screen, test compounds were evaluated at asingle concentration (10 �M). As an assessment of screenrobustness, Z� values for each plate were calculated; only plateswith Z� � 0.5 were evaluated further. Percentage inhibition foreach compound was determined, and a hit cut-off for the pri-mary screen was set at mean � 1 � S.D. of all results (�40%inhibition); hits were confirmed by retesting in triplicate in theprimary assay at 10 �M. Confirmed hits were further evaluatedin 10-point, half-log concentration response assays.

Kinome Screening Panel—Compounds were screened at aconcentration of 1 �M in competition binding assays at AmbitBiosciences as described previously (27). The technologyemploys kinases either produced as fusions to T7 phage orexpressed as fusions to NF-�B in HEK-293 cells and subse-quently tagged with DNA for PCR detection. Competition withtest compound for binding to resin-conjugated affinity ligands,as measured by quantitative PCR, was used to determine thepotency of compounds against each kinase.

Chemogenomics—Using a collection of 21,851 compoundswith kinome-wide profiles against the panel of 386 kinases (17),compounds were excluded from the chemogenomics analysis ifany of the following criteria were met: observed JAK2 or JAK3activity, defined as �35% control in Ambit ligand displacementassays (or IC50 � 0.5 �M in internal enzymatic assays); �20kinases inhibited in the kinase panel with potency �67% inhi-bition at 1 �M; or �67% inhibition against any of CDK2, CDK5,or CDK7 in the Ambit ligand displacement assays. Only 7,908of the 21,314 compounds passed all of these criteria.

Compounds—Synthesis of Compound 1 ((R)-N-(1-(3-(8-methyl-5-(methylamino)-8H-imidazo[4,5-d]thiazolo[5,4-b]pyridin-2-yl)phenyl)ethyl)-2-(methylsulfonyl)benzamide)was as follows. A solution of 2-bromo-5-chloro-8-methyl-8H-imidazo[4,5-d]thiazolo[5,4-b]pyridine (200 mg, 0.659mmol, prepared as described in Ref. 18) and (R)-tert-butyl1-(3-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)phenyl)-ethylcarbamate (297 mg, 0.856 mmol, prepared as describedin Ref. 19) in N,N-dimethylformamide (2 ml), EtOH(1.2 ml), and water (0.4 ml) was degassed by sparging with astream of nitrogen gas for 5 min. To this mixture was thenadded PdCl2(1,1�-bis(diphenylphosphino)ferrocene) (48.2mg, 0.066 mmol) and sodium carbonate (349 mg, 3.29mmol). After sparging with nitrogen gas for another 5 min,the reaction vessel was sealed and heated at 100 °C for 5 h.After cooling, the reaction mixture was filtered through asyringe tip filter (PTFE) and purified by preparative reversephase HPLC to give the product (R)-tert-butyl 1-(3-(5-chloro-8-methyl-8H-imidazo[4,5-d]thiazolo[5,4-b]pyridin-2-yl)phenyl)ethylcarbamate (150 mg, 51% yield) as a solid.This product (80 mg, 0.18 mmol) was added to a 2.0 M solu-tion of methanamine (0.90 ml, 1.80 mmol) in water to give a

Tyk2 Pseudokinase Domain Stabilizers Block Signal Transduction

11062 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 290 • NUMBER 17 • APRIL 24, 2015

by guest on Novem

ber 6, 2020http://w

ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

suspension that was heated by microwave at 150 °C for 30min. The resulting dark brown solution was filtered througha PTFE filter and purified by reverse phase preparative HPLCto give (R)-tert-butyl 1-(3-(8-methyl-5-(methylamino)-8H-imidazo[4,5-d]thiazolo[5,4-b]pyridin-2-yl)phenyl)ethylcar-bamate (75 mg, 76% yield) as a white solid. Hydrochloric acid(2 M in EtOH, 0.8 �l, 1.6 mmol) was then added in one por-tion to the product, and the resulting heterogeneous mixturewas stirred at room temperature for 12 h before removingthe solvent under reduced pressure. The residue was furtherdried under vacuum to give (R)-2-(3-(1-aminoethyl)phenyl)-N,8-dimethyl-8H-imidazo[4,5-d]thiazolo[5,4-b]pyridin-5-amine (50 mg, 93% yield). To a solution of this product fromthe last step (50 mg, 0.148 mmol) in N,N-dimethylformamide(1 ml) was added (1H-benzotriazol-1yloxy)[tris(dimethyl-amino)]phosphonium hexafluorophosphate (78 mg, 0.177mmol), diisopropyl ethylamine (0.077 ml, 0.443 mmol), and2-(methylsulfonyl)benzoic acid (44.4 mg, 0.222 mmol). Theresulting mixture was stirred at room temperature for 5 h. Afterpassing through a syringe tip PTFE filter, the reaction mixturewas purified by reverse phase preparative HPLC to give Com-pound 1 (22 mg, 23% yield) as a white powder. 1H NMR (400MHz, CDCl3) � ppm 8.08 (1 H, d), 7.95 (1 H, d, J � 2.0 Hz), 7.86(1 H, d, J � 7.7 Hz), 7.60 –7.78 (4 H, m), 7.51 (2 H, t, J � 7.7 Hz),5.45 (1 H, q), 4.32 (3 H, s), 3.37–3.45 (3 H, m), 3.28 (3 H, br s),1.75 (3 H, d, J � 6.8 Hz). MS (E�) m/z: 521 (MH�).

[3H]Compound 1 ((R)-N-(1-(3-(8-methyl-5-(methylamino)-8H-imidazo[4,5-d]thiazolo[5,4-b]pyridin-2-yl)phenyl)ethyl)-2-([3H]methylsulfonyl)benzamide) was prepared as follows.2-Mercaptobenzoic acid (2.3 mg, 0.015 mmol) and cesium car-bonate (2 mg, 0.006 mmol) were added to a 5-ml round-bot-tomed flask. The flask was attached to a ported glass vacuumline, and anhydrous N,N-dimethylformamide (0.5 ml) wasintroduced with magnetic stirring. An ampoule of tritiatedmethyl iodide (200 mCi, PerkinElmer Life Sciences lot3643419) was added to the reaction flask, and stirring wasmaintained at room temperature for 3 h. Without purification,the crude product was reacted with 3-chloroperoxybenzoicacid (10 mg, 0.058 mmol) predissolved in CH2Cl2 (1 ml) at roomtemperature with stirring. The reaction was stirred for 7 h, andadditional 3-chloroperoxybenzoic acid (10 mg, 0.058 mmol)was added. The reaction was stirred for �24 h, and HPLC anal-ysis indicated 35– 40% conversion to the desired sulfone prod-uct. The crude product was purified by semipreparative HPLCto give 81 mCi of 2-([3H]methylsulfonyl)benzoic acid productidentified by its HPLC co-elution with an authentic standard.The product was dissolved in anhydrous acetonitrile to give afinal solution activity of 5.8 mCi/ml. Ten ml of the solution (58mCi) were transferred into a 25-ml round-bottomed flask,which was then attached to a vacuum line, and the solution wascarefully evaporated to dryness. (R)-2-(3-(1-Aminoethyl)phe-nyl)-N,8-dimethyl-8H-imidazo[4,5-d]thiazolo[5,4-b]pyridin-5-amine (prepared as described above for the synthesis ofCompound 1, 1.1 mg, 0.0033 mmol) and benzotriazol-1-yl-oxytripyrrolidinophosphonium hexafluorophosphate (2 mg,0.0053 mmol) dissolved in anhydrous N,N-dimethylformamide(1.5 ml) were added to the flask, followed by N,N-diisopropyl-ethylamine (0.010 ml). The resulting clear solution was stirred

at room temperature for 4 h. The crude reaction mixture waspurified by semipreparative HPLC to yield a total of 18 mCi ofthe desired product in 99.9% radiochemical purity. Mass spec-tral analysis of the tritiated product (m/z M � H 527.33) wasused to establish the specific activity at 81 Ci/mmol.

The synthesis of BMS-066 has been reported previously (20).MLN120B was obtained from ChemScene (Monmouth Junc-tion, NJ).

Signaling Assays in kit225 Cells and Peripheral Blood TCells—STAT3 phosphorylation in kit225 cells induced by 25ng/ml IL-23 (R&D Systems, 1290) was determined by ELISA(Cell Signaling, 7146) after a 20-min stimulation followed bythe addition of lysis buffer (Cell Signaling) containing proteaseand phosphatase inhibitors. Activation of Tyk2 was determinedin the same cells by measuring phosphorylation of Tyr-1054and Tyr-1055 using an antibody that specifically recognizesTyk2 phosphorylated at these sites (BD Biosciences). Briefly,cells were stimulated with 50 ng/ml IL-23 for 10 min, lysed,analyzed by SDS-PAGE (10% BisTris 1.0-mm gel, NuPageWG1201), and transferred to nitrocellulose membranes.Immunoblotting and detection with AlexaFluor680 goat anti-mouse IgG (Invitrogen, A21057) was performed, and the blotswere analyzed using the LI-COR Odyssey scanner. The quanti-fication of phosphorylated Tyk2 was normalized to actin, blot-ted simultaneously.

The effect of Compound 1 on IFN�- or IL-15-induced STATphosphorylation was measured in human peripheral bloodmononuclear cells stimulated with either 1000 units/ml IFN�(Peprotech, Rocky Hill, NJ) or 25 ng/ml IL-15 (Peprotech) for15 min. Reactions were stopped by BD Phosflow lyse/fix buffer(BD Biosciences) at 37 °C, and after staining with CD3 fluores-cein isothiocyanate-conjugated antibody, pSTAT1 PerCp-Cy5.5 (Tyr(P)-701), pSTAT3 PE (Tyr(P)-705), and pSTAT5(Tyr(P)694 Alexa647-conjugated) antibodies, flow cytometrywas used to gate on CD3� T cells, and the amount of STATphosphorylation was determined. Samples were acquired on aFACSCanto II using DIVA 6.1.1 software (BD Biosciences) andanalyzed using FlowJo Vx (Tree Star). All conjugated antibodieswere from BD Biosciences. Measurements of thrombopoietin-induced phosphorylation of STAT5 in platelets from humanperipheral blood were performed as described previously (21).

Expression and Purification of Tyk2 Pseudokinase Domain—The coding region of the human Tyk2 pseudokinase domain(residues 575– 869) was generated by PCR and cloned as anNdeI-XhoI fragment into a modified pFastBac1 vector (Invit-rogen) with an N-terminal His-TVMV tag to generate His-TVMV-hTyk2 pseudokinase domain (residues 575– 869)-pFB.The DNA sequence of the PCR product was sequence-verified.The TVMV cleavage sequence is closely related to the TEVcleavage site, and the TVMV viral protease recognizes theseven-residue sequence (ETVRFQ2G) with high selectivity,which leaves a single additional residue at the N terminus of thecleaved protein.

Baculovirus was generated for the His-TVMV-hTyk2 pseu-dokinase domain (residues 575– 869) construct using the Bac-to-Bac baculovirus expression system (Invitrogen) according tothe manufacturer’s protocol. For large scale protein produc-tion, Sf9 cells (Expression Systems, Davis, CA) grown in ESF921



by guest on Novem

ber 6, 2020http://w

ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

insect medium (Expression Systems) at 2 � 106 cells/ml wereinfected with virus stock at a 1:100 virus/cell ratio for 66 h. Theproduction was carried out at 20-liter scale in a 50-liter Cellbagusing the WAVE-Bioreactor system 20/50 (GE Healthcare).The infected cells were harvested by centrifugation, and wetinsect cell pellet was dissolved at a 2.5:1 mass ratio of Buffer A(50 mM Hepes, pH 7.7, 500 mM NaCl, 25 mM imidazole, 5% (v/v)glycerol, 0.1% Triton X-100, and 0.5 mM TCEP), supplementedwith protease inhibitor tablets (Roche Applied Science, catalogno. 13514100) and Benzonase (Sigma, catalog no. E8263-25KU). The cells were lysed by sonication, and the lysate wasclarified by centrifugation at 9,500 rpm (Thermo, F10S-6x500Yrotor) for 30 min at 4 °C. Purification steps were executed usingan AKTA Explorer-100 system. The supernatant was applied toa nickel affinity column (5-ml HisTrap FF crude, GE Health-care, code 17-5286-01) washed to baseline with 50 mM Hepes,pH 7.7, 500 mM NaCl, 25 mM imidazole, 5% (v/v) glycerol, and0.5 mM TCEP and then eluted with 50 mM Hepes, pH 7.7, 500mM NaCl, 350 mM Imidazole, 5% (v/v) glycerol, and 0.5 mM

TCEP. Fractions containing the target protein were concen-trated with Amicon Ultra-15 centrifuge units (Millipore, cata-log no. UFC901096). After concentration, the protein was fur-ther purified by size exclusion chromatography (HiLoadSuperdex 75 (16/60), GE Healthcare, code 17-1068-01) run in50 mM Hepes, pH 7.7, 500 mM NaCl, 1 mM MnCl2, 5% (v/v)glycerol, 0.5 mM TCEP. For biophysical and crystallizationexperiments, the N-terminal His tag was removed by overnightcleavage at 4 °C, using His-tobacco vein mottling virus proteaseat a 10:1 mass ratio. The next day, the His-tobacco vein mot-tling virus protease and cleaved hexahistidine tag were boundto a 5-ml HisTrap FF crude column while the processed Tyk2pseudokinase domain (residues 575– 869) protein passedthrough the column. The protein was then buffer-exchangedinto either 50 mM Na2HPO4, pH 7.7, 500 mM NaCl, 5% (v/v)glycerol, and 5 mM DTT) for long term storage at 80 °C fol-lowing flash freezing or alternatively 10 mM Na2HPO4, pH 7.7,500 mM NaCl, 5% (v/v) glycerol, and 5 mM DTT for complexingwith ligand and crystallization trials, using a HiPrep 26/10desalting column (GE Healthcare, catalog no. 17-5087-01). Theaverage yield was 5 mg/liter of insect cell culture. The purifiedprotein was analyzed by SDS-PAGE, dynamic light scattering(Wyatt DynaPro Plate reader), and liquid chromatography-mass spectrometry to establish purity, determine the associa-tion state, and confirm the mass, respectively.

Binding and Biophysical Assays—Thermal shift analysis(TSA) was used to probe small molecule binding to the hTyk2pseudokinase domain (residues 575– 869) using a Bio-Rad PCRinstrument (CFX96/C1000 Thermocycler; 96-well plate for-mat) and SYPRO� Orange indicator dye. Samples were pre-pared to a final volume of 10 �l each, in triplicate, using a ther-mal scan profile of 30 –90 °C at a scan rate of 60 °C/h. Finalconditions in each well were the following: 0.15 mg/ml (4.25�M) protein in buffer containing 50 mM Na2HPO4, pH 7.7, 500mM NaCl, 5% (v/v) glycerol, 5 mM DTT, 2 mM MgCl2, 5% (v/v)DMSO-d6 containing Compound 1 at concentrations spanning0 –33.3 �M, and 1� SYPRO� Orange (diluted from a DMSOstock). Working compound concentrations were based on aDMSO-d6 stock sample of the molecule, in which the concen-

tration was established by NMR quantitation relative to aninternal standard. Data were analyzed with customized soft-ware, fitting baselines and the midpoint of the thermal transi-tion, which was reported as the Tm value (°C), with mean andS.D. values from triplicates reported.

Binding of human Tyk2 pseudokinase domain (residues575– 869) to the small molecule ligand, Compound 1, was eval-uated by isothermal titration calorimetry using a VP-ITCinstrument (MicroCal/GEHC) thermally controlled at 25 °C.The protein was extensively dialyzed against fresh buffer (10mM Na2HPO4, pH 7.7, 500 mM NaCl) at 4 °C prior to titration,and the dialysate was retained for preparing the small moleculesolution. A stock solution of Compound 1 was prepared to �10mM in DMSO-d6, and the concentration was measured byNMR and reference to an internal standard. A working solutionof �50 �M was prepared, by diluting the stock solution in thebuffer dialysate to a final DMSO-d6 concentration of 0.50%(v/v). Protein concentration was adjusted by dilution to �5 �M

and confirmed by absorbance readings based on the amino acidcomposition (�280 � 31,160 M1 cm1). A small matching vol-ume (i.e. 0.50% (v/v)) of DMSO-d6 was added to the samplebefore transferring into the sample cell to minimize DMSOsample mismatch. Titrations were carried out in triplicate, andcontrol injections of buffer into protein and Compound 1 intoblank buffer were performed, both of which showed minimalbackground heats. A stirring rate of 307 rpm was maintained inthe experiments, with an initial 2-�l injection, followed by15–24 injections that varied between 5 and 10 �l in volume,spaced every 180 s. These injection profiles produced a 1.5–3.0-fold excess of ligand/protein for complete definition of thebinding isotherm. The manufacturer’s software was used toanalyze and plot the data, fitting to a 1:1 binding model.

A binding assay measuring the binding of [3H]Compound 1to the N-terminal His-tagged Tyk2 pseudokinase domain wasdeveloped. Assays were performed in 384-well plates with afinal assay volume of 20 �l containing copper-polyvinyltoluenescintillation proximity assay beads (PerkinElmer Life Sciences,catalog no. RPNQ0095) at 80 �g/ml, [3H]Compound 1 (20 nM),the N-terminal His-tagged Tyk2 pseudokinase domain (2.5nM), and test compounds in assay buffer (50 mM HEPES, pH 7.5,100 �g/ml BSA, 5% DMSO). After incubating at room tem-perature for 30 min, the inhibition was calculated by thedisplacement of [3H]Compound 1 binding as determined byscintillation counting. Dose-response curves were generatedto determine the concentration required to inhibit [3H]-Compound 1 binding by 50% (IC50).

Crystallization, Data Collection, and Structure Determinationof Tyk2 Pseudokinase Domain in Complex with BMS-066—The Tyk2 pseudokinase domain (residues 575– 869) at 1.2mg/ml in 10 mM sodium phosphate, pH 7.7, 500 mM NaCl, 5%glycerol, 5 mM DTT was complexed with a 3-fold molar excessof compound, incubated overnight at 4 °C, and concentrated to2.3 mg/ml. Crystals were grown using the hanging drop vapordiffusion method in a drop consisting of 1 �l of protein solutionand 1 �l of reservoir solution containing 30% PEG 5000 (methylether), 200 mM ammonium sulfate, and 100 mM sodium caco-dylate buffer, pH 6.5, on Qiagen EasyXtal plates (Qiagen Sci-ences, Inc.). Macroseeding was performed to obtain diffraction



by guest on Novem

ber 6, 2020http://w

ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

size crystals. The crystals, grown at room temperature, began toappear in 1 week and took 2–3 weeks to reach full size (0.1 �0.1 � 0.15 mm). Crystals were flash-cooled in liquid nitrogenwith 25% glycerol and 75% reservoir solution as cryoprotectant.A 1.80 Å resolution data set was collected by Shamrock Struc-tures, Inc. at Canadian Light Source beamline 08ID-1 with aRayonix MX-300 detector. The data were processed and scaledusing the program HKL2000 (22). The crystals had symmetryconsistent with space group P1 with cell dimensions a � 41.5 Å,b � 63.5 Å, c � 68.0 Å (� � 74.4°, � 75.5°, � � 88.2°) and twomolecules in the crystallographic asymmetric unit. The struc-ture was determined by molecular replacement using the pro-gram Phaser (23) with an Abl kinase starting model (ProteinData Bank code 2GQG), where the residue side chains werestubbed, the activation loop and P-loop were removed, and themodel was split into N-terminal (residues 243–319) and C-ter-minal (residues 320 –500) domains. For molecular replace-ment, the two C-terminal domain orientations were found first,followed by the two N-terminal domains. Iterative cycles ofmodel building and refinement using the programs refmac (24),autoBUSTER (25), and Coot (26) were carried out to build inside chain atoms and the missing loops. The model at this pointwas sufficient to clearly identify electron density for BMS-066.The location of the ligand was at the interface of the N- andC-terminal domains, near the hinge at residue Tyr-689 for bothTyk2 pseudokinase domain molecules. The final model of theTyk2 pseudokinase domain consists of residues 580 – 609,636 –785, and 792– 869 in molecule A and residues 580 – 609,636 –785, and 792– 867 in molecule B (Tyk2 pseudokinasedomain numbering), two ligand molecules, 25 sulfate ions, and257 water molecules.

RESULTS

Chemogenomics Approach to Kinases Involved in IL-23R Sig-nal Transduction—A library of 7,908 compounds selected froma collection of kinase inhibitors synthesized over the course ofvarious kinase-targeted drug discovery efforts at Bristol-MyersSquibb throughout the last decade was used in a chemogenom-ics approach to identify inhibitors of IL-23R signal transductionin cells and elucidate their target mechanisms. The library waschosen for evaluation from a larger collection for which broadkinase inhibitory profiles were obtained at an inhibitor concen-tration of 1 �M against a panel of 386 unique human kinase andpseudokinase domains at Ambit Biosciences (hereafter referredto as the kinome screening panel), which employs competitionbinding assays to determine inhibitor potency (27). Becausekinase inhibitors with broad inhibition across the kinomewould confound the interpretation of the subsequent analysisof kinases involved in IL-23 signaling, compounds that were notreasonably selective (i.e. bind to more than 20 of 386 kinases inthe panel) or likely to be simply cytotoxic to cells (e.g. cyclin-de-pendent kinase inhibitors) were excluded from the chemog-enomics analysis. In addition, because JAK2 and Tyk2 werepreviously understood to be involved in IL-23R signal transduc-tion, compounds were also excluded if they were known inhib-itors of the enzymatic activity of these JAK family kinases.

Using kit225 human T cells into which a firefly luciferasereporter gene under the control of the interferon-� activation

sequence had been stably transfected, this library was evaluatedat a concentration of 10 �M against IL-23-stimulated expres-sion of the luciferase reporter gene. Hits showing �40% inhibi-tion in the initial screen were confirmed by retesting in tripli-cate, with most confirmed hits evaluated in concentrationresponse assays to calculate IC50 values.

With these cellular screening results, each kinase in thescreening panel was evaluated for the “cell active rate” (CAR)defined as the percentage of compounds inhibiting a particularkinase by more than 90% at 1 �M that also inhibited the IL-23-stimulated reporter assay in kit225 T cells with an IC50 of �1�M (or �75% inhibition at 10 �M if no IC50 had been gener-ated). Associating cell activity with the percentage of inhibitorsagainst a particular kinase allows for the potential identificationof kinases regulating the cellular end point, even when usingcompounds that may not be exquisitely selective, and normal-izes for the number of inhibitors of that particular kinase withinthe collection of compounds tested. A particular kinase is notnecessarily important in regulating the IL-23 pathway based onthe results with a promiscuous kinase inhibitor. Identifyinginhibitors of that particular kinase with high cell active rateacross the known inhibitors of that kinase, however, provides amore confident approach to identifying kinases truly involvedin the pathway, but still requires additional follow up validation.As shown in Fig. 1A and supplemental Table S1, inhibitors ofonly three kinases in the screening panel yielded a correspond-ing CAR of greater than 10%. These include inhibitors of GSK-3�, IKK, and the pseudokinase (JH2) domain of Tyk2, whichshowed CAR values of 10.5, 12.2, and 15.6%, respectively. Theidentification of GSK-3� is not particularly surprising, becausea role for GSK3 kinases in the regulation of STAT activation hasbeen reported (28), although GSK3 kinases have not previouslybeen implicated in IL-23R signal transduction specifically. Thefinding of a high CAR value for IKK, a critical regulator ofcanonical NF-�B activation, was not expected because the pro-moter upstream of the luciferase reporter in kit225 cells doesnot contain an NF-�B response element. Subsequent examina-tion of the IKK inhibitors active against IL-23-induced lucif-erase expression in kit225 T cells revealed that the compoundsrepresented imidazothiazolopyridines and related IKK inhib-itors that also bind the pseudokinase domain of Tyk2. In otherwords, the apparent CAR for IKK and the Tyk2 pseudokinasedomain result from a common set of compounds that bind toboth proteins. Potent IKK inhibitors unrelated to this chemi-cal series, such as MLN120B (29), do not bind to the Tyk2 pseu-dokinase domain and are inactive in the IL-23-stimulatedkit225 T cell assay. These observations, along with additionalevaluations on the mechanism of action (see below), indicatethat the cellular activity of this set of compounds results fromtheir action on the Tyk2 pseudokinase domain.

A representative example of this class of Tyk2 pseudokinasedomain binders is Compound 1 (Fig. 1B) which, as shown in arepresentation of the kinome-wide profiling in Fig. 1C, is verypotent against the Tyk2 pseudokinase (99% inhibition at 1 �M)but lacks potency against all other kinases in the kinome assaypanel except IKK (96% inhibition) and the JAK1 pseudokinasedomain (99% inhibition at 1 �M). No evidence of binding to thecatalytic domain of Tyk2 or any other JAK family kinase was



by guest on Novem

ber 6, 2020http://w

ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

evident at 1 �M, and subsequent enzymatic assays confirmedthe lack of activity against purified catalytic domains (IC50 �� 2�M). The pseudokinase domains of JAK2 and JAK3 were notpart of the kinome screening panel. In the IL-23-stimulatedkit225 T cell assay, Compound 1 inhibited the stimulatedresponse with an IC50 of 485 143 nM (n � 3).

Biophysical Characterization of the Binding of Compound 1to the Pseudokinase Domain of Tyk2—The ability of Compound1 to bind to the Tyk2 pseudokinase domain, as indicated by thekinome screening panel assay, was confirmed by isothermaltitration calorimetry (ITC) and TSA. These experiments uti-lized a purified recombinant human protein comprising resi-dues 575– 869 of the Tyk2 pseudokinase domain. As shown inFig. 2A, ITC measurements established that a high affinitybinding interaction resulted from binding of the Tyk2 pseu-

dokinase domain and Compound 1, with a KD value in the lownanomolar range. The mean value obtained from triplicateexperiments was KD � 5 4 nM, approaching the limit of deter-mination under the direct titration method used here. Theexperiments revealed an exothermic binding reaction, with amean value of �Hobs � 9.4 0.2 kcal/mol, reflecting favor-able changes in binding enthalpy for the system, a value withina typical range for many small molecule-protein interactions(30). Because the titration was performed in phosphate buffer,which has �Hion

0 of �0 kcal/mol, the observed enthalpy, �Hobs,approaches �Hbiol for this system. In other words, the observedenthalpy does not reflect significant changes from buffer pro-tonation/deprotonation but instead reflects net changes in thebound versus free states of protein and ligand (31). Conse-quently, the majority of the free energy change (�G) of this

FIGURE 1. Identification of Tyk2 pseudokinase domain binders as inhibitors of IL-23R signal transduction. A, CAR for 386 kinases in the kinome screeningpanel. CAR is defined as the percentage of compounds inhibiting a particular kinase by more than 90% at 1 �M that also inhibited the IL-23-stimulated reporterassay in kit225 T cells (see “Results” for details). B, chemical structures of compounds detailed in the present work. C, kinase inhibitory profile of Compound 1at 1 �M across 386 kinases in the kinome screening panel, with kinases inhibited by �99% (red circles), �90% (yellow circles), or �90% (blue circles). This figureis reproduced courtesy of Cell Signaling Technology, Inc. D, inhibition by Compound 1 of IL-23-induced luciferase reporter expression (gray circles) andphosphorylation of STAT3 (blue squares) in kit225 T cells. The results represent the average and S.D. (error bars) of triplicate measurements (*, p � 0.05).



by guest on Novem

ber 6, 2020http://w

ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

binding reaction reflects exothermic stabilization (�H) of thesystem. The method also yielded a stoichiometry for the inter-action between Compound 1 and the Tyk2 pseudokinasedomain of 0.5– 0.7, modestly lower than 1.0. The data wereinterpreted as reflecting 1:1 binding, consistent with the x-raystructure (see below).

To determine the effect of Compound 1 binding on the Tyk2pseudokinase domain, the thermal unfolding of protein wasmonitored by fluorescence spectroscopic measurements withSYPRO Orange. The fluorescence emission of this dye signifi-cantly increases as hydrophobic surfaces of globular proteinsare exposed to the aqueous solution upon heat-induced dena-turation. As shown in Fig. 2B, the apo-Tyk2 pseudokinasedomain showed a midpoint on the protein stability curve (melt-ing temperature, Tm) of 43.5 °C, and titration of Compound 1with a 1–2-fold molar excess resulted in an increase in the Tm to�50 °C, with a maximum �Tm � 14 °C at 8-fold more excess.This large positive �Tm shift demonstrated that Compound 1was bound to a region of the protein that manifests an increasein stability, although this method cannot be used to determinea KD value (32). Combining the results from both the ITC andTSA experiments, Compound 1 shows enthalpically driven,high affinity binding to the Tyk2 pseudokinase domain, whichresults in a pronounced stabilization of the pseudokinasedomain. Recombinant, full-length Tyk2 is poorly behaved andcould not be used in thermal shift assays.

Pharmacologic Stabilization of the Tyk2 Pseudokinase Inhib-its Activation and STAT Signaling—Because the pseudokinasedomain is thought to lack catalytic activity, precluding the useof an enzymatic activity assay to measure compound potency, ahomogeneous ligand displacement assay was developed to eval-uate the potency of additional analogs in concentrationresponse determinations. Using a radiolabeled version of Com-pound 1 ([3H]Compound 1) as the ligand in a homogeneousbinding assay using the Tyk2 pseudokinase domain complexedto scintillation proximity assay beads, a KD of �10 nM was

determined. As detailed in Fig. 3A, evaluation of compounds inthis series as well as in chemically unrelated compoundsshowed that a strong correlation exists between potency againstthe IL-23-stimulated kit225 T cell assay and in the Tyk2 pseu-dokinase domain, as measured by displacement of [3H]Com-pound 1 (Spearman’s correlation coefficient � 0.8). Asexpected, there was no correlation between cellular and IKKpotency (see Fig. 3B). A limitation of the present work is that aninhibitor-insensitive mutant form of Tyk2 is not available tofurther demonstrate that these compounds act through theTyk2 pseudokinase domain.

In an effort to probe the mechanism by which Tyk2 pseu-dokinase domain stabilizers, such as Compound 1, inhibitIL-23-induced transcriptional responses, the effect on Tyk2-catalyzed phosphorylation of STAT3 was determined. AsshowninFig.1D,Compound1 inhibitedIL-23-stimulatedphos-phorylation of STAT3 in kit225 T cells with a potency equipo-tent to that obtained against transcription of the STAT-respon-sive reporter gene in the same experiment, indicating that theTyk2 pseudokinase domain stabilizer blocks Tyk2-catalyzedphosphorylation of STAT proteins even without directly inter-acting with the catalytic domain of Tyk2. Because phosphory-lation of Tyr-1054 and Tyr-1055 within the activation loop ofthe catalytic domain of Tyk2 is critical in inducing a catalyti-cally active form of the kinase (33), the effect of Compound 1 onTyr-1054/Tyr-1055 phosphorylation upon receptor-mediatedstimulation was investigated. Fig. 3C shows that Compound 1inhibited the IL-23-stimulated phosphorylation of Tyr-1054/Tyr-1055 in a concentration-dependent manner, with apotency (IC50 �500 nM) consistent with the effects on bothSTAT3 phosphorylation and STAT-dependent reporter geneexpression in these cells.

Many of the cytokine and growth factor receptors that relyupon JAK kinases often utilize two members of the family tomediate downstream signal transduction, and JAK2 alongwith Tyk2 has been implicated in IL-23R signaling. In order

FIGURE 2. Biophysical characterization of the binding of Compound 1 to the Tyk2 pseudokinase domain. A, ITC data showing titration of Compound 1into Tyk2 pseudokinase domain (residues 575– 869) at 25 °C. A representative example of three replicate experiments is shown, with the top panel representingthe time course of the injection profile. Each negative peak represents injection of a small volume of ligand into the cell solution, containing Tyk2 pseudokinasedomain at 5.0 �M. The bottom panel shows the integrated heats per injection as closed circles plotted as a function of molar equivalents of ligand/protein. Thesolid line in the bottom panel represents the fit to the data, using a 1:1 binding model. B, mean Tm values for the Tyk2 pseudokinase domain (residues 575– 869)titrated with Compound 1, as determined by TSA with SYPRO� Orange. The open circles represent the mean Tm value observed from triplicate determinationsof the thermal unfolding of the protein in the presence of varying concentrations of Compound 1 (in �M). The error bars reflect the S.D. (error bars) of the Tmvalues, with all points shown being statistically significant compared with no compound (p � 0.05).



by guest on Novem

ber 6, 2020http://w

ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

to determine the specificity for Tyk2-dependent functionalsignaling, Compound 1 was evaluated in human peripheralblood T cells for effects on the phosphorylation of STATproteins using agonists of receptors both dependent on andindependent of Tyk2. Type I interferon receptor signalingrequires both Tyk2 and JAK1, and as shown in Fig. 3D, Com-pound 1 inhibited IFN�-induced phosphorylation ofSTAT1, STAT3, and STAT5 with equivalent potency (IC50�500 nM). However, the compound was ineffective againstsignaling pathways that do not require Tyk2, as evidenced bythe lack of activity against both thrombopoietin-stimulatedphosphorylation of STAT5 (JAK2-dependent signaling) andIL-15-induced phosphorylation of STAT5 (JAK1- and JAK3-dependent signaling).

Because Compound 1 at 1 �M showed inhibition against thepseudokinase domain of JAK1 in the kinome screening panelnearly equal to that shown against the Tyk2 pseudokinasedomain, the lack of functional activity against JAK1-dependentIL-15 receptor signaling is surprising. The results may suggestthat the mechanism by which the pseudokinase domain ofJAK1 regulates the catalytic domain differs from Tyk2 andthat JAK1 pseudokinase domain binders cannot prevent the

activation of kinase activity. Alternatively, it may be that theisolated recombinant JAK1 pseudokinase domain can obtaina conformation in vitro that accommodates binding of Com-pound 1, but the endogenous full-length multidomain pro-tein present in cells cannot. Although we do not have wellbehaved full-length JAK1 (or Tyk2) recombinant proteins tomore fully discriminate between these potential explana-tions for the observed difference in functional activity, it isimportant to note that more than 100 other compounds thatappear to be JAK1 pseudokinase binders have been evaluatedand failed to show functional potency in JAK1-dependentcellular assays (data not shown). This indicates that theeffect is not unique to Compound 1. The pseudokinasedomains of JAK2 and JAK3 have not been evaluated to deter-mine whether Compound 1 can bind in vitro, but our datasuggest that, at least for this class of Tyk2 pseudokinasebinders, there is a high level of functional selectivity in cel-lular assays. As detailed below, the Tyk2 pseudokinasedomain binding site for this class of compounds has uniquefeatures when compared with the pseudokinase domains ofthe other JAK family members, which may provide somerationale for the observed selectivity.

FIGURE 3. Cellular characterization of the potency and mechanism of Tyk2 pseudokinase domain stabilizers. A, correlation between potency againstTyk2 pseudokinase domain potency as measured by displacement of [3H]Compound 1 versus the IL-23-induced reporter assay in kit225 T cells. B, correlationbetween potency in an IKK enzymatic assay versus the IL-23-induced reporter assay in kit225 T cells for the same set of compounds. C, inhibition by Compound1 of IL-23-induced phosphorylation of Tyr-1054/Tyr-1055 of Tyk2 in kit225 cells. D, effect of Compound 1 on ligand-induced phosphorylation of STAT proteinsin peripheral blood T cells as measured by flow cytometry. Open squares, IFN�-induced pSTAT1; black triangles, IFN�-induced pSTAT3; gray circles, IFN�-induced pSTAT5; open diamonds, thrombopoietin-induced pSTAT5; black circles, IL-15-pSTAT5. The results are the average of three donors measured induplicate (*, p � 0.05). Error bars, S.D.



by guest on Novem

ber 6, 2020http://w

ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

Crystal Structure of Tyk2 Pseudokinase Domain with BoundStabilizer—To help understand the mechanism by which Com-pound 1 and related Tyk2 pseudokinase domain stabilizers pre-vent the activation of the catalytic domain and to aid futurecompound design efforts, we obtained crystal structures of theTyk2 pseudokinase domain bound with compounds in thisseries. We were unsuccessful in obtaining a crystal structure ofTyk2 pseudokinase domain in complex with Compound 1;however, an x-ray crystal structure bound to a less potent ana-log, BMS-066 (see Fig. 1B), was determined to 1.8 Å resolution.This compound showed IC50 values of 72 and 1020 nM againstthe Tyk2 pseudokinase domain probe displacement and IL-23-stimulated reporter assays, respectively (data not shown). Asshown in Fig. 4A, the crystal structure of the Tyk2 pseudokinase

domain encompasses the main structural features of a proteinkinase. Namely, the protein fold is separated into two sub-domains, or lobes, with the smaller N-terminal lobe composedof a five-stranded sheet and an additional short strand fromresidues 589 –591 in the N terminus as well as the prominenthelix �-C. The C-terminal lobe is larger than the N-terminallobe and is predominantly helical. The two lobes are connectedthrough a single polypeptide strand (the linker/hinge region),which in catalytic kinases is normally involved in binding tothe adenine ring of ATP. The site analogous to the ATP-binding site in catalytic kinases forms a deep cleft betweenthe two lobes and sits beneath the phosphate binding loop(P-loop) connecting strands 1 and 2. BMS-066 is observedto bind in this cleft.

FIGURE 4. Crystal structure of the Tyk2 pseudokinase domain with bound BMS-066. A, ribbon representation of protein with BMS-066 shown in stickrepresentation and green carbons. B, Tyk2 pseudokinase domain residues corresponding to those of protein kinases normally involved in catalytic machineryare highlighted. C, P-loop interactions with activation loop (hydrogen bonds shown as dashed lines) highlighting the only phosphorylatable residue inactivation loop, Ser-768, and a bound sulfate ion. D, surface representation of Tyk2 pseudokinase domain structure. The region highlighted within the dashedoval depicts the concave surface of the protein near the activation loop. Protein carbons are in green, oxygens in red, and nitrogens in blue. BMS-066 withmagenta carbons is shown in a stick representation.



by guest on Novem

ber 6, 2020http://w

ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

Consistent with this domain being defined as a pseudokinase,the crystal structure shows a number of amino acid and struc-tural differences that, when compared with catalytic kinases,indicate that this domain is unable to support catalytic activity.For example, the typically conserved 3-strand lysine (Lys-642)is present but does not form a salt bridge with a conservedglutamate from helix �-C (important for maintaining an activeconformation in catalytic kinases). Instead, Thr-658 occupiesthe position of the typical conserved glutamate. The precedingresidue in the helix is interestingly Glu-657, but it points towardsolvent (see Fig. 4B). Lys-642 in the pseudokinase domainstructure is instead near hydrogen bonding distance to Asp-759, the aspartate that in catalytic kinases belongs to the con-served DFG motif, which marks the N-terminal portion of theactivation loop and normally interacts with the ATP phos-phates through Mg2�. In the Tyk2 pseudokinase domain, thismotif is replaced by DPG, whereby proline appears to restrictthe position of Asp-759 to a region normally occupied by thehelix �-C glutamate. The altered position of Lys-642 and Asp-759 would not optimally interact with Mg2�-ATP. Moreover,the P-loop that normally serves as a gate to ATP binding incatalytic protein kinases usually contains a conserved glycine-rich sequence motif (GXGXG), where , usually tyrosine orphenylalanine, caps the site of phosphate transfer, but in theTyk2 pseudokinase domain the sequence is GXGXRT, whichmight be expected to reduce the flexibility of the P-loop due tothe loss of the glycine and the rigidifying effects of the hydrogenbonding interactions of the P-loop Arg-600 (see below). A sul-fate ion (a crystallization solution ingredient) is observed underthe P-loop and is also involved in hydrogen bonds to the sidechains of the P-loop Thr-599 and Thr-601 and the backboneNH of Thr-599. The activation loop, which in catalytic kinasesnot only provides the binding platform for peptide substrateinteractions but also ensures the appropriate orientation of thecatalytic machinery, is truncated compared with the catalyticdomain activation loop (20 residues versus 26) and adopts apartially helical conformation. Interactions that appear tostitch the N-terminal lobe and the C-terminal lobe together tohinder ATP and substrate binding include a salt bridge betweenthe P-loop Arg-600 and the activation loop Glu-771 (both con-served only in the JAK1 pseudokinase domain) and a hydrogenbond between Arg-738 and the P-loop carbonyl of Gln-597through a water (Fig. 4C). Perhaps the most important differ-ence compared with catalytic kinases is that the canonical pro-tein kinase catalytic loop motif HRD (D is the catalytic aspartatethat contributes to the nucleophilicity of the substratehydroxyl) is replaced by HGN (Asn-734) in the Tyk2 pseudoki-nase domain. Asn-734 is observed to be involved in a hydrogenbond with the backbone carbonyl of Pro-760 of the DPG motif(Fig. 4C). Interestingly, the surface of the protein near the acti-vation loop, usually the site of substrate binding, is rather con-cave in nature (see Fig. 4D), which may indicate that it could beinvolved in protein-protein interactions. In summary, the par-ticular conformation of the activation loop, the interactionsbetween the P-loop and C-terminal lobe (particularly the acti-vation loop), and the absence of the catalytic aspartate do notappear to be consistent with an ability of the domain to bind

either ATP or peptide substrate and catalyze phosphoryl trans-fer as an active kinase.

BMS-066 sits in the deep cleft between the N- and C-termi-nal lobes, with the tricyclic ring system of BMS-066 occupyingthe site typically bound by the adenine group of ATP in catalytickinases (see Fig. 5A). The imidazole basic nitrogen and theamino moiety form hydrogen bonds with Val-690 of the hinge.The imidazole N-methyl group sits in a hydrophobic pocketcomposed of 3-strand Val-640 and 6-strand Leu-741, andthe pendant pyridine ring is sandwiched between the 1-strandof the P-loop and Pro-694 and Arg-738 of the extended hinge.The amide NH hydrogen of the methoxyacetamide grouphydrogen bonds with the backbone carbonyl of the P-loop Leu-595 as the group protrudes out toward solvent. A water mole-cule deep in the pocket forms hydrogen bonds with the sidechain of the gatekeeper Thr-687 and the hinge carbonyl fromGlu-688, and there is also a water hydrogen bond with the pyr-idine nitrogen of the tricycle. Docking of Compound 1 into thesolved Tyk2 pseudokinase domain structure suggests that thehinge-binding interactions are identical to BMS-066 but thatthe rest of the molecule may bind deeper into the binding site,burying more surface area and possibly making a number ofadditional hydrogen bonds, which might explain its enhancedpotency. A potential binding model is provided in Fig. 6.

Recently, structures of the Tyk2 pseudokinase domain andthe Tyk2 pseudokinase-kinase dual domain were released in theProtein Data Bank (PDB entries 3ZON and 4OLI, respectively(34). An �-carbon backbone alignment of the structuredescribed here and the released versions reveals that the Tyk2pseudokinase domain structures adopt very similar conforma-tions (PDB entry 3ZON, root mean square deviation 0.7 Å; PDBentry 4OLI, root mean square deviation 0.8 Å). Comparison ofthe overall fold of the Tyk2 pseudokinase domain structurewith recently published JAK1 and JAK2 JH2 structures (35, 36)also shows high similarity (Fig. 5B).

Not only are there differences in the Tyk2 pseudokinasedomain from catalytic kinases related to its inability to functionas a kinase, but a sequence comparison of the BMS-066 bindingsite residues with the 491 sequences of the kinome provides thebasis for high selectivity of a Tyk2 pseudokinase domain stabi-lizer. Besides the difference in orientation of the conserved Lys-642 and Asp-759 of DPG compared with a protein kinase in theactive state, there are a number of other uncommon residuesnear the BMS-066 binding site. As shown in Fig. 5C, these dif-ferences include Pro-760 of the DPG motif, which is not presentwithin any other kinase. In addition, the gatekeeper residue is aThr (Thr-687, 19% prevalence in kinome), the residue belowthe gatekeeper is an Ala (Ala-671, 2% in kinome), and the resi-due before DPG is a Ser (Ser-758, 15% kinome), with the com-bination of the three being unique to the kinome and indicatingthat the pocket is well suited to design highly selective thera-peutics. In the extended hinge, Pro-694 is found in only about2% of the kinome, and the nearby Val-697 is found in less than1% of the kinome. In addition, the binding site residues withinthe JAK JH2 family are not highly conserved, presenting anopportunity to design pseudokinase domain selectivity as well(Fig. 7A).



by guest on Novem

ber 6, 2020http://w

ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

DISCUSSION

Typically, kinase inhibitors have been targeted to the ATPbinding site, and although this approach has been successful inidentifying potent inhibitors, the design of therapeutic agentswith selectivity across a target class of more than 500 humanprotein kinases can be quite challenging due to the highly con-served nature of this binding site. Selectivity has proven to beespecially difficult to obtain within the JAK family kinases,which have highly conserved active sites. Targeting the pseu-

dokinase domain of Tyk2, as identified using the chemogenom-ics strategy detailed in the present report, represents a novelapproach in the design of highly selective inhibitors of Tyk2-dependent signaling downstream of the IL-23 receptor andother pathways important in autoimmunity. Although it hadbeen previously understood that the pseudokinase domains ofTyk2 and the other JAK family kinases play a critical role in theregulation of receptor-mediated activation of the catalyticdomain, Compound 1 and related imidazothiazolopyridines

FIGURE 5. Binding of BMS-066 to Tyk2 pseudokinase domain. A, interactions of BMS-066 (carbons colored in magenta, oxygens in red, nitrogens in blue) withTyk2 pseudokinase domain (carbons in green); B, superposition of Tyk2 pseudokinase domain/BMS-066 structure (green) with JAK1 pseudokinase domain(blue; PDB code 4L00) and JAK2 JH2 (magenta; PDB code 4FVQ) to show similar fold; C, Tyk2 pseudokinase domain structure bound with BMS-066, highlightingresidues that are compared with representation in kinome. Percentages are calculated as number of kinases that possess a given residue out of 491 kinases.



by guest on Novem

ber 6, 2020http://w

ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

represent the first demonstration that this mechanism can betargeted by small molecule agents to block receptor-mediatedsignaling. That pseudokinase stabilizers block the activation ofthe kinase domain is especially relevant to autoimmunitybecause a coding variant of Tyk2 resulting in a change fromPro-1104 to alanine has been shown to similarly block activa-tion of the catalytic domain (37), and individuals carrying thisallele are protected from multiple sclerosis and possibly otherautoimmune disorders (37, 38).

Although the exact molecular mechanism by which the pseu-dokinase domain of Tyk2 regulates receptor-mediated activa-tion of the catalytic domain is not known, much of the pub-lished data support a model where the pseudokinase domainnormally forms an intramolecular interaction with the catalyticdomain such that the pseudokinase domain either induces aninactive conformation of the catalytic domain, or the pseudoki-nase domain blocks access of substrates or ATP by a directinteraction in this region (34, 36, 39, 40). A putative receptor-mediated conformational change in the pseudokinase domainwould then break the intramolecular autoinhibitory interactionwith the catalytic domain, allowing the catalytic domain toachieve a conformation that allows for both phosphorylation ofthe activation loop and full catalytic activity. Although a recentreport has suggested that the JAK2 JH2 (pseudokinase) domainmay have catalytic activity and can autophosphorylate bothSer-523 and Tyr-570 autoinhibitory phosphorylation siteswithin the catalytic domain (35), it is unlikely that this proposedmechanism is relevant to Tyk2. Not only are the phosphoryla-tion sites implicated in JAK2 not conserved in Tyk2, but thestructural features detailed above argue that the Tyk2 pseu-dokinase domain can neither accommodate binding of ATPnor catalyze phosphoryl transfer. Indeed, even in comparingthe crystal structures of the Tyk2 and JAK2 JH2/pseudokinasedomains, there are key differences in residues that line the ATPbinding pocket. Lys-677 of the JAK2 JH2 domain, engaged inhydrogen bonds with the ATP phosphates, is an arginine in theTyk2 pseudokinase domain and adopts a different conforma-tion, pointing away from the binding site. In addition, the ribosering of ATP binds near Ser-633 of the JAK2 JH2, but the corre-

sponding residue in Tyk2 is Pro-694, which would appear tosterically clash with the JAK2-bound ATP. We have beenunable, furthermore, to demonstrate that ATP or non-hydro-lyzable ATP analogs can bind to the Tyk2 pseudokinasedomain, either through the use of biophysical methods, such asthermal shift assays, or by displacement of radiolabeled Com-pound 1 binding to the Tyk2 pseudokinase domain (results notshown).

Detailed work with JAK1 is consistent with the pseudokinasedomain forming an inhibitory interaction with the catalyticdomain (36), and a model of this intramolecular autoinhibitoryinteraction between pseudokinase domain and catalyticdomains of JAK kinases postulates that one of the contact sur-faces includes an �-C helix-�-C helix interface, and anotherone includes the catalytic domain activation loop, where thelatter would be precluded from adopting an active conforma-tion (41). Additional evidence of an intramolecular autoinhibi-tory interaction mediated by the pseudokinase domain of Tyk2was recently provided by a crystal structure of a dual domainpseudokinase-kinase construct, with the interaction appearingto limit the conformational mobility of the catalytic domainactive site (34). Upon a receptor-induced conformationalchange in the pseudokinase domain mediated via the FERMdomain and SH2-pseudokinase (SH2-PK) linker, the intramo-lecular autoinhibitory interaction with the catalytic domain isbroken, allowing the catalytic domain to achieve a conforma-tion that allows for both phosphorylation of the activation loopand catalytic activity. Studies of the 5-strand Val to Phe muta-tions in the JH2/pseudokinase domains of JAK1, JAK2, andTyk2, all of which lead to constitutive activity (34, 42, 43, 44),are consistent with this model. In the case of JAK1, structuralanalysis of the JAK1 pseudokinase domain WT and V658Fmutant domains showed the mutant V658F packs in an edge-to-face manner with �-C helix Phe-636 and fills in the sameposition occupied by Phe-575 of the SH2-PK linker in the wild-type structure. Therefore, the V658F mutation induces a coor-dinated rearrangement of the SH2-PK linker. Interestingly, oneof the monomers in the asymmetric unit of the wild type JAK1pseudokinase domain structure also displays this rearrange-ment, which has been proposed to be a structural switch con-trolling catalytic activation. The residues involved in re-arrangement in JAK1 and JAK2 are conserved in the Tyk2pseudokinase domain except for �-C helix Phe (JAK2 Phe-595),which is conservatively replaced by a Tyr (Fig. 7B). Based on ourcurrent understanding, the mechanism of catalytic domain reg-ulation by the pseudokinase domain involving �-C helix andthe SH2-PK linker interactions in JAK1 are probably similar forTyk2. It is therefore both surprising and intriguing that Com-pound 1 is selective for receptor-mediated pathways dependenton Tyk2 compared with those receptors dependent on otherJAK family members because the compound binds to the pseu-dokinase domain of JAK1 but does not impact JAK1-dependent(but Tyk2-independent) receptor signaling. The compoundalso failed to block JAK2- and JAK3-dependent signaling. Theselectivity for Tyk2-dependent signaling may reflect a mecha-nistic attribute of Tyk2 that is unique, or the apparent bindingto the JAK1 pseudokinase domain is artifactual. An increased

FIGURE 6. A potential binding model of Compound 1 in the Tyk2 pseu-dokinase domain. Interactions of Compound 1 (carbons colored in magenta,oxygens in red, nitrogens in blue) with Tyk2 pseudokinase domain (carbons ingreen). Hydrogen bonds are noted by dashed lines.



by guest on Novem

ber 6, 2020http://w

ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

understanding of JAK family kinase regulation will be requiredto fully account for these observations.

In summary, the present results are consistent with a modelin which small molecule ligands of the Tyk2 pseudokinasedomain act to lock it in a conformation that stabilizes an auto-inhibitory interaction with the catalytic domain. This stabiliza-tion prevents receptor-mediated activation and/or catalyticactivity of the catalytic domain, blocking downstream signaltransduction and STAT-dependent gene transcription. Target-ing the pseudokinase domain of Tyk2 represents a novelapproach in the design of highly selective therapeutics targetingTyk2-dependent signaling important in autoimmunity. Whileadditional investigations into both the mechanisms by whichTyk2 pseudokinase domain stabilizers block the activation ofthe catalytic domain and the unique attributes of Tyk2 com-pared with other JAK family members are under way, we arecapitalizing on this finding to optimize potent and selectiveagents for the treatment of human autoimmune diseases.Detailed evaluations of the pharmacology of selective Tyk2pseudokinase stabilizers will be the subject of a future report.

Acknowledgments—We thank Greg Ford for providing the IL-23-re-sponsive luciferase reporter cells; Joann Strnad and Mian Gao forinput into the experimental design; and Hua Gong, Scott Watterson,and Alaric Dyckman for contributions to the synthesis of Compound 1and BMS-066.

REFERENCES1. Miossec P., and Kolls, J.K. (2012) Targeting IL-17 and TH17 cells in

chronic inflammation. Nat. Rev. Drug Discov. 11, 763–7762. Volpe, E., Servant, N., Zollinger, R., Bogiatzi, S. I., Hupe, P., Barillot, E., and

Soumelis, V. (2008) A critical function for transforming growth factor-,interleukin 23 and proinflammatory cytokines in driving and modulatinghuman T(H)-17 responses. Nat. Immunol. 9, 650 – 657

3. Geremia, A., Arancibia-Carcamo, C. V., Fleming, M. P., Rust, N., Singh, B.,Mortensen, N. J., Travis, S. P., and Powrie, F. (2011) IL-23-responsive

innate lymphoid cells are increased in inflammatory bowel disease. J. Exp.Med. 208, 1127–1133

4. Sandborn, W. J., Gasink, C., Gao, L. L., Blank, M. A., Johanns, J., Guzzo, C.,Sands, B. E., Hanauer, S. B., Targan, S., Rutgeerts, P., Ghosh, S., de Villiers,W. J., Panaccione, R., Greenberg, G., Schreiber, S., Lichtiger, S., Feagan,B. G., and CERTIFI Study Group (2012) Ustekinumab induction andmaintenance therapy in refractory Crohn’s disease. N. Engl. J. Med. 367,1519 –1528

5. Strober, B. E., Crowley, J. J., Yamauchi, P. S., Olds, M., and Williams D. A.(2011) Efficacy and safety results from a phase III, randomized controlledtrial comparing the safety and efficacy of briakinumab with etanercept andplacebo in patients with moderate to severe chronic plaque psoriasis. Br. J.Dermatol. 165, 661– 668

6. Longbrake, E. E., and Racke, M. K. (2009) Why did IL-12/IL-23 antibodytherapy fail in multiple sclerosis? Expert Rev. Neurother. 9, 319 –321

7. Vollmer, T. L., Wynn, D. R., Alam, M. S., and Valdes, J. (2011) A phase 2,24-week, randomized, placebo-controlled, double-blind study examiningthe efficacy and safety of an anti-interleukin-12 and -23 monoclonal anti-body in patients with relapsing-remitting or secondary progressive mul-tiple sclerosis. Mult. Scler. 17, 181–191

8. Murray, P. J. (2007) The JAK-STAT signaling pathway: input and outputintegration. J. Immunol. 178, 2623–2629

9. Vincenti, F., Tedesco Silva, H., Busque, S., O’Connell, P., Friedewald, J.,Cibrik, D., Budde, K., Yoshida, A., Cohney, S., Weimar, W., Kim, Y. S.,Lawendy, N., Lan, S. P., Kudlacz, E., Krishnaswami, S., and Chan, G. (2012)Randomized phase 2b trial of tofacitinib (CP-690,550) in de novo kidneytransplant patients: efficacy, renal function and safety at 1 year. Am. J.Transplant. 12, 2446 –2456

10. Oyamada, A., Ikebe, H., Itsumi, M., Saiwai, H., Okada, S., Shimoda, K.,Iwakura, Y., Nakayama, K. I., Iwamoto, Y., Yoshikai, Y., and Yamada, H.(2009) Tyrosine kinase 2 plays critical roles in the pathogenic CD4 T cellresponses for the development of experimental autoimmune encephalo-myelitis. J. Immunol. 183, 7539 –7546

11. Ortmann, R., Smeltz, R., Yap, G., Sher, A., and Shevach, E. M. (2001) Aheritable defect in IL-12 signaling in B10.Q/J mice. I. In vitro analysis.J. Immunol. 166, 5712–5719

12. Liang, J., van Abbema, A., Balazs, M., Barrett, K., Berezhkovsky, L., Blair,W., Chang, C., Delarosa, D., DeVoss, J., Driscoll, J., Eigenbrot, C., Ghilardi,N., Gibbons, P., Halladay, J., Johnson, A., Kohli, P. B., Lai, Y., Liu, Y.,Lyssikatos, J., Mantik, P., Menghrajani, K., Murray, J., Peng, I., Sambrone,A., Shia, S., Shin, Y., Smith, J., Sohn, S., Tsui, V., Ultsch, M., Wu, L. C., Xiao,

FIGURE 7. Unique features of pseudokinase domain of Tyk2 compared with other JAK members. A, Tyk2 pseudokinase domain structure bound withBMS-066 highlighting residues that are compared with a representation in the JAK pseudokinase domain family. B, superposition of Tyk2 pseudokinasedomain/BMS-066 structure (green) with JAK1 pseudokinase domain (magenta; PDB code 4L00) and JAK2 JH2 (yellow; PDB code 4FVQ), highlighting residuesthat may be involved in stabilization and triggering release of the autoinhibitory hold of the pseudokinase domain over the kinase domain.



by guest on Novem

ber 6, 2020http://w

ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

Y., Yang, W., Young, J., Zhang, B., Zhu, B. Y., and Magnuson, S. (2013)Lead optimization of a 4-aminopyridine benzamide scaffold to identifypotent, selective, and orally bioavailable TYK2 inhibitors. J. Med. Chem.56, 4521– 4536

13. Norman, P. (2012) Selective JAK1 inhibitor and selective Tyk2 inhibitorpatents. Expert Opin. Ther. Pat. 22, 1233–1249

14. Zeqiraj E., and van Aalten, D.M. (2010) Pseudokinases-remnants of evo-lution or key allosteric regulators? Curr. Opin. Struct. Biol. 20, 772–781

15. Saharinen, P., Vihinen, M., and Silvennoinen, O. (2003) Autoinhibition ofJak2 tyrosine kinase is dependent on specific regions in its pseudokinasedomain. Mol. Biol. Cell. 14, 1448 –1459

16. Saharinen, P., and Silvennoinen, O. (2002) The pseudokinase domain isrequired for suppression of basal activity of Jak2 and Jak3 tyrosine kinasesand for cytokine-inducible activation of signal transduction. J. Biol. Chem.277, 47954 – 47963

17. Posy, S. L., Hermsmeier, M. A., Vaccaro, W., Ott, K. H., Todderud, G.,Lippy, J. S., Trainor, G. L., Loughney, D. A., Johnson, S. R. (2011) Trends inkinase selectivity: insights for target class-focused library screening.J. Med. Chem. 54, 54 – 66

18. Kempson, J., Spergel, S. H., Guo, J., Quesnelle, C., Gill, P., Belanger, D.,Dyckman, A. J., Li, T., Watterson, S. H., Langevine, C. M., Das, J., Moquin,R. V., Furch, J. A., Marinier, A., Dodier, M., Martel, A., Nirschl, D., VanKirk, K., Burke, J. R., Pattoli, M. A., Gillooly, K., McIntyre, K. W., Chen, L.,Yang, Z., Marathe, P. H., Wang-Iverson, D., Dodd, J. H., McKinnon, M.,Barrish, J. C., and Pitts, W. J. (2009) Novel tricyclic inhibitors of I�B kinase.J. Med. Chem. 52, 1994 –2005

19. Dyckman, A., Pitts, W. J., Belema, M., Gill, P., Kempson, J., Qiu, Y.,Quesnelle, C., Spergel, S. H., and Zusi, F. C. (November 25, 2008) Imidazo-fused Oxazolo[4,5-B]pyridine and Imidazo-fused Thiazolo[4,5-B]pyri-dine Based Tricyclic Compounds and Pharmaceutical CompositionsComprising Same, United States Patent 7456194

20. Gillooly, K. M., Pattoli, M. A., Taylor, T. L., Chen, L., Cheng, L., Gregor,K. R., Whitney, G. S., Susulic, V., Watterson, S. H., Kempson, J., Pitts, W. J.,Booth-Lute, H., Yang, G., Davies, P., Kukral, D. W., Strnad, J., McIntyre,K. W., Darienzo, C. J., Salter-Cid, L., Yang, Z., Wang-Iverson, D. B., andBurke, J. R. (2009) Periodic, partial inhibition of I�B kinase -mediatedsignaling yields therapeutic benefit in preclinical models of rheumatoidarthritis. J. Pharmacol. Exp. Ther. 331, 349 –360

21. Purandare, A. V., McDevitt, T. M., Wan, H., You, D., Penhallow, B., Han,X., Vuppugalla, R., Zhang, Y., Ruepp, S. U., Trainor, G. L., Lombardo, L.,Pedicord, D., Gottardis, M. M., Ross-Macdonald, P., de Silva, H., Hosbach,J., Emanuel, S. L., Blat, Y., Fitzpatrick, E., Taylor, T. L., McIntyre, K. W.,Michaud, E., Mulligan, C., Lee, F. Y., Woolfson, A., Lasho, T. L., Pardanani,A., Tefferi, A., Lorenzi, M. V. (2012) Characterization of BMS-911543, afunctionally selective small-molecule inhibitor of JAK2. Leukemia 26,280 –288

22. Otwinowski, Z., and Minor, W. Processing of x-ray diffraction data col-lected in oscillation mode (1987) Methods Enzymol. 276, 307–326

23. McCoy, A. J., Grosse-Kunstleve, R. W., Adams, P. D., Winn, M. D., Sto-roni, L. C., and Read, R. J. (2007) Phaser crystallographic software. J. Appl.Crystallogr. 40, 658 – 674

24. Murshudov, G. N., Skubak, P., Lebedev, A. A., Pannu, N. S., Steiner, R. A.,Nicholls, R. A., Winn, M. D., Long, F., and Vagin, A. A. (2011) REFMAC5for the refinement of macromolecular crystal structures. Acta Crystallogr.D Biol. Crystallogr. 67, 355–367

25. Bricogne, G., Blanc, E., Brandl, M., Flensburg, C., Keller, P., Paciorek, W.,Roversi, P., Sharff, A., Smart, O. S., Vonrhein, C., and Womack, T. O.(2011) BUSTER, Global Phasing Ltd., Cambridge, UK

26. Emsley, P., Lohkamp, B., Scott, W.G., and Cowtan, K. (2010) Features andDevelopment of Coot. Acta Crystallogr. D Biol. Crystallogr. 66, 486 –501

27. Davis, M. I., Hunt, J. P., Herrgard S., Ciceri, P., Wodicka, L. M., Pallares, G.,Hocker, M., Treiber, D. K., Zarrinkar, P. P. (2011) Comprehensive analysisof kinase inhibitor selectivity. Nat. Biotechnol. 29, 1046 –1051

28. Beurel, E., and Jope, R. S. (2008) Differential regulation of STAT familymembers by glycogen synthase kinase-3. J. Biol. Chem. 283, 21934 –21944

29. Nagashima, K., Sasseville, V. G., Wen, D., Bielecki, A., Yang, H., Simpson,

C., Grant, E., Hepperle, M., Harriman, G., Jaffee, B., Ocain, T., Xu, Y., andFraser, C. C. (2006) Rapid TNFR1-dependent lymphocyte depletion invivo with a selective chemical inhibitor of IKK. Blood 107, 4266 – 4273

30. Freire, E. (2008) Do enthaply and entropy distinguish first in class frombest in class? Drug Discov. Today 13, 869 – 874

31. Doyle M. L., Louie, G., Dal Monte P. R., and Sokoloski, T. D. (1995) Tightbinding affinities determined from thermodynamic linkage to protons bytitration calorimetry. Methods Enzymol. 259, 183–194

32. Matulis, D., Kranz, J. K., Salemme, F. R., and Todd, M. J. (2005) Thermo-dynamic stability of carbonic anhydrase: measurements of binding affinityand stoichiometry using ThermoFluor. Biochemistry 44, 5258 –5266

33. Gauzzi, M. C., Velazquez, L., McKendry, R., Mogensen, K. E., Fellous, M.,and Pellegrini, S. (1996) Interferon-�-dependent activation of Tyk2 re-quires phosphorylation of positive regulatory tyrosines by another kinase.J. Biol. Chem. 271, 20494 –20500

34. Lupardus, P. J., Ultsch, M., Wallweber, H., Bir Kohli, P., Johnson, A. R., andEigenbrot, C. (2014) Structure of the pseudokinase-kinase domains fromprotein kinase TYK2 reveals a mechanism for Janus kinase (JAK) autoin-hibition. Proc. Natl. Acad. Sci. U.S.A. 111, 8025– 8030

35. Bandaranayake, R. M., Ungureanu, D., Shan, Y., Shaw, D. E., Silvennoinen,O., Hubbard, S. R. (2012) Crystal structures of the JAK2 pseudokinasedomain and the pathogenic mutant V617F. Nat. Struct. Mol. Biol. 19,754 –759

36. Toms, A., Deshpande, A., McNally, R., Jeong, Y., Rogers, J. M., Kim, C. U.,Gruner, S. M., Ficarro, S. B., Marto, J. A., Sattler, M., Griffin, J. D., and Eck,M. J. (2013) Structure of a pseudokinase-domain switch that controlsoncogenic activation of Jak kinases. Nat. Struct. Mol. Biol. 20, 1221–1223

37. Couturier, N., Bucciarelli, F., Nurtdinov, R. N., Debouverie, M., Lebrun-Frenay, C., Defer, G., Moreau, T., Confavreux, C., Vukusic, S., Cournu-Rebeix, I., Goertsches, R. H., Zettl, U. K., Comabella, M., Montalban, X.,Rieckmann, P., Weber, F., Müller-Myhsok, B., Edan, G., Fontaine, B.,Mars, L. T., Saoudi, A., Oksenberg, J. R., Clanet, M., Liblau, R. S., andBrassat, D. (2011) Tyrosine kinase 2 variant influences T lymphocyte po-larization and multiple sclerosis susceptibility. Brain 134, 693–703

38. Ban, M., Goris, A., Lorentzen, A. R., Baker, A., Mihalova, T., Ingram, G.,Booth, D. R., Heard, R. N., Stewart, G. J., Bogaert, E., Dubois, B., Harbo,H. F., Celius, E. G., Spurkland, A., Strange, R., Hawkins, C., Robertson,N. P., Dudbridge, F., Wason, J., De Jager, P. L., Hafler, D., Rioux, J. D.,Ivinson, A. J., McCauley, J. L., Pericak-Vance, M., Oksenberg, J. R., Hauser,S. L., Sexton, D., Haines, J., Sawcer, S., Wellcome Trust Case-ControlConsortium (WTCCC), and Compston, A. (2009) Replication analysisidentifies TYK2 as a multiple sclerosis susceptibility factor. Eur. J. Hum.Genet. 17, 1309 –1313

39. Percy, M.J., and McMullin, M.F. (2005) The V617F JAK2 mutation and themyeloproliferative disorders. Hematol. Oncol. 23, 91–93

40. Constantinescu S.N., Leroy, E., Gryshkova, V., Pecquet, C., and Dusa, A.(2013) Activating Janus kinase pseudokinase domain mutations in my-eloproliferative and other blood cancers. Biochem. Soc. Trans. 41, 1048 –1054

41. Giordanetto, F., and Kroemer, R.T. (2002) Prediction of the structure ofhuman Janus kinase 2 (JAK2) comprising JAK homology domains 1through 7. Protein Eng. 15, 727–737

42. Pieri, L., Bogani, C., Guglielmelli, P., Zingariello, M., Rana, R. A., Barta-lucci, N., Bosi, A., and Vannucchi, A. M. (2009) The JAK2V617 mutationinduces constitutive activation and agonist hypersensitivity in basophilsfrom patients with polycythemia vera. Haematologica 94, 1537–1545

43. Staerk, J., Kallin, A., Demoulin, J. B., Vainchenker, W., and Constanti-nescu, S. N. (2005) JAK1 and Tyk2 activation by the homologous polycy-themia vera JAK2 V617F mutation: cross-talk with IGF1 receptor. J. Biol.Chem. 280, 41893– 41899

44. Shide, K., Shimoda, K., Kamezaki, K., Kakumitsu, H., Kumano, T., Nu-mata, A., Ishikawa, F., Takenaka, K., Yamamoto, K., Matsuda, T., andHarada, M. (2007) Tyk2 mutation homologous to V617F Jak2 is not foundin essential thrombocythaemia, although it induces constitutive signalingand growth factor independence. Leuk. Res. 31, 1077–1084



by guest on Novem

ber 6, 2020http://w

ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

Elkin, Yuval Blat, William J. Pitts, David S. Weinstein and James R. BurkeJohnson, Sophie Wu, Suzanne C. Edavettal, Yang Hong, Mark R. Witmer, Lisa L.

Chang, Dianlin Xie, Lihong Cheng, Donna Pedicord, Jodi Muckelbauer, Stephen R. John S. Tokarski, Adriana Zupa-Fernandez, Jeffrey A. Tredup, Kristen Pike, ChiehYing

Pharmacological Stabilization of Its Pseudokinase DomainTyrosine Kinase 2-mediated Signal Transduction in T Lymphocytes Is Blocked by

doi: 10.1074/jbc.M114.619502 originally published online March 11, 20152015, 290:11061-11074.J. Biol. Chem.

10.1074/jbc.M114.619502Access the most updated version of this article at doi:

Alerts:

When a correction for this article is posted•

When this article is cited•

to choose from all of JBC's e-mail alertsClick here

Supplemental material:

http://www.jbc.org/content/suppl/2015/03/11/M114.619502.DC1

http://www.jbc.org/content/290/17/11061.full.html#ref-list-1

This article cites 41 references, 14 of which can be accessed free at

by guest on Novem

ber 6, 2020http://w

ww

.jbc.org/D

ownloaded from

http://www.jbc.org/lookup/doi/10.1074/jbc.M114.619502

http://www.jbc.org/cgi/alerts?alertType=citedby&addAlert=cited_by&cited_by_criteria_resid=jbc;290/17/11061&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/290/17/11061

http://www.jbc.org/cgi/alerts?alertType=correction&addAlert=correction&correction_criteria_value=290/17/11061&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/290/17/11061

http://www.jbc.org/cgi/alerts/etoc

http://www.jbc.org/content/suppl/2015/03/11/M114.619502.DC1

http://www.jbc.org/content/290/17/11061.full.html#ref-list-1

http://www.jbc.org/

tyrosinekinase2-mediatedsignaltransductionint ... · a crystal structure of the pseudoki-nase...

Documents